Numerous studies have examined factors that contribute to racial and ethnic disparities in health outcomes for newborns. An issue that has not been well studied is variation in the use of neonatal intensive care units (NICUs) and variation in the use of high quality hospitals. Studies in other areas (e.g. cardiac care) document large racial and ethnic disparities in treatment use and use of high quality providers, raising questions about the role that this might play in newborn outcome disparities as well. We will study racial and ethnic variation in use of NICUs, particularly the use of advanced and high volume NICUs, and use of high quality hospitals by newborns, particularly LBW and VLBW newborns, using secondary datasets that link vital statistics and hospital discharge data for all births in 7 states over most of the 1990s. Our analysis of high quality hospital use will be based on new measures of hospital quality based on patient outcomes that we will develop using new and innovative methods. Our main analysis is designed to achieve 3 specific aims: 1) to investigate the extent of differences in use of NICUs and high volume facilities by race and ethnicity, and examine the correlates of differences in utilization rates; 2) to construct new outcome-based indicators of the quality of hospital care for newborns, and examine variations in the quality of hospitals at which newborns of different races and ethnicities are born or to which they are transferred, and to examine the correlates of differences; and 3) to examine the effect of differences in access to NICUs, high volume facilities, and high quality hospitals on disparities in health outcomes for newborns by race and ethnicity. The datasets we will use also afford more extensive opportunities to control for co-morbidities and other risk factors than have been available in previous large-scale analyses of newborn outcome disparities, as well as presenting the opportunity to study the role played by geographic ("small area") variations in newborn outcome disparities. As we conduct our analysis of disparities in NICU and high quality hospital use, we will also apply these capabilities of the new datasets to the study of disparities, working to achieve specific aim 4: to use the new linked databases to extend work on racial and ethnic disparities in health outcomes for newborns, particularly accounting for variations in co-morbidities and other risk factors, newly available in the new linked datasets, and studying effects of variation across geographic areas. Our work will focus on Black-White differences, disparities for Hispanics (including subgroups like Mexican and Puerto Rican), and disparities for Asian and Pacific Islanders. [unreadable] [unreadable]